I know this is not so much a webmin or virtualmin thing, but on my box that normally runs them, I got into serious troubles today and I'm hoping the solution, or at least the approach of solving this problem is just a normal logical step for one of you (while I just don't know...). Here it is:
Hi,
my server is not booting right and i'm in troubles now...
it's a 6.06 ubuntu server (don't ask). It hasn't been rebooted for quite some months, but today it hung and I had to reboot it.
It's at fasthosts.co.uk, and I've got some 'vkm' (java) remote control for it.
Rebooting didn't 'start' the server as it should, as no sites were accessible. None of the services were accessible, not ssh, not ftp etc etc
with the vkm remote controle i could see what was going on, I get a console and see the server as if it's local.
The last thing it does when starting up, is showing that it's started Postfix [ok] and then nothing.
If I hit enter then, I can login. Internally, it then seems all fine and services are up and running.
webmin wasn't, but was easy to start. With 'links' I can see it works on https://127.0.0.1:10101.
However, it can't ping outside or see external websites. From outside, nothing can see the server either.
My questions: How do I find out what the server is supposed to be doing after loading Postfix, as that's where the problem might be?
What could cause this communication problem?
Iptables is empty at this stage, no rules
If I type: ping http://www.google.com it says: unknown host http://www.google.com
It's unable to resolve anything outside. Looks like a DNS issue then, doesn't it?
netstat -nr gives empty result.
If I type ifconfig, i get this at inet6 addr:
inet6 addr: ::1/128 Scope:Host
as if something's missing, no?
And then there's this:
ifup -a
Does anybody have an idea what I should do next?
Thanks!
(almost panicking for not having this server up and running...)
You could be seeing a networking or DNS issue...
Two things you'd want to do --
First, run an "ifconfig", and make sure eth0 shows up with your external IP address.
Second, verify that /etc/resolv.conf exists, and has the correct nameservers in it. You'd also want to make sure your BIND daemon is online.
Those two are a good place to start, we can figure things out from there :-)
-Eric
Thanks Eric!
I ran ifconfig and the card didn't show up.
When I restarted networking services (/etc/init.d/networking restart) things got back to normal.
BIND was online, and resolv.conf existed and the contents were ok.
Now I have to try to find out why it doesn't boot normally, but that's not as urgent.
Thanks again.
Using
ifconfig -ayou can, if the error repeats, check if the eth device is basically present and just not assigned correctly. I had this happen when the MAC address of a virtual eth changed./etc/network/interfacesholds the intended configuration for your eths, and/etc/udev/rules.d/persistent-net.rulesshould contain a list of PCI network devices with required MAC addresses and which eth they will become. (Yeah, it's all a bit more complex under Linux... but you can do nearly everything if you know where to look. ;) )Additionally, the logfiles in
/var/log, especiallymessages,dmesgandsyslog, might contain information about errors during startup.Note: I'm using Ubuntu 10.04, so some of this information might not yet apply to your "merry old" distro. ;)